Closed chuck-ma closed 2 days ago
Hi, thanks for your work. But I'm wondering how to create an int4 model like svdq-int4-flux.1-dev.safetensors, because I have a model fine-tuned from the dev model.
Gotcha, it seems that i should use deepcompressor to create w4a4 quantized model
Hi, thanks for your work. But I'm wondering how to create an int4 model like svdq-int4-flux.1-dev.safetensors, because I have a model fine-tuned from the dev model.